ARTICLES AND REVIEWS
Blood Lead Levels in Mexico and Pediatric Burden
of Disease Implications
Jack Caravanos, DrPH, CIH, Russell Dowling, MPH, Martha María Téllez-Rojo, Dra,
Alejandra Cantoral, ScD, Roni Kobrosly, PhD, MPH, Daniel Estrada,
Manuela Orjuela, MD, ScM, Sandra Gualtero, MSc, Bret Ericson, MSc,
Anthony Rivera, and Richard Fuller
ABSTRACT
Background: Although there has been success in reducing lead exposure with the phase-out of leaded gasoline, exposure to
lead in Mexico continues to threaten the health of millions, much of which is from lead-based glazes used in pottery that leaches
into food.
Objectives: An extensive historical review and analysis of available data on blood lead levels in Mexican populations was
conducted. We used a calculated geometric mean to evaluate the effect of lead on the pediatric burden of disease.
Methods: An extensive bibliographic search identified 83 published articles from 1978 to 2010 with blood lead level (BLL)
data in Mexican populations representing 150 data points from more than 50,000 study participants. Values from these
publications were categorized into various groupings. We then calculated the incidence of disease and disability-adjusted life-years
resulting from these BLLs using the World Health Organization’s burden of disease spreadsheets for mild mental retardation.
Results: Reviewing all relevant studies, the geometric means of Mexican BLLs in urban and rural areas were found to be 8.85 and
22.24 ug/dL, respectively. Since the phase-out of leaded gasoline, the mean in urban areas was found to be 5.36 ug/dL and the average
in rural areas is expected to be much higher. The U.S. Centers for Disease Control and Prevention’s (CDC) upper limit of blood lead in
children under the age of 6 years is 5 ug/dL and the current U.S. average is 1.2 ug/dL. Our results indicate that more than 15% of the
population will experience a decrement of more than 5 IQ points from lead exposure. The analysis also leads us to believe that lead is
responsible for 820,000 disability-adjusted life-years for lead-induced mild mental retardation for children aged 0 to 4 years.
Conclusion: Lead continues to threaten the health of millions and remains a significant cause of disability in Mexico.
Additional interventions in reducing or managing lead-based ceramic glazes are necessary to protect the public health.
Key Words: lead, blood lead level, burden of disease, children, DALY, disability-adjusted life-year, pottery, Mexico
2014 Icahn School of Medicine at Mount Sinai. Annals of Global Health 2014;80:269-277
INTRODUCTION
2214-9996/ª 2014 Icahn School of Medicine at Mount Sinai
From the Environmental and Occupational Health Sciences Program, City
University of New York, School of Public Health, New York, NY (JC);
Blacksmith Institute, New York, NY (JC, RD, RK DE, SG, BE, AR, RF);
Instituto Nacional de Salud Pública, Universidad No. 655 Colonia Santa
María Ahuacatitlán, Cerrada Los Pinos y Caminera C.P. Cuernavaca,
Morelos, México (MMT-R, AC); Department of Preventive Medicine, Icahn
School of Medicine at Mount Sinai, New York, NY (RK); Departments of
Pediatrics and Environmental Health Sciences, Columbia University, New
York, NY (MO). Address correspondence to J.C.; e-mail: jcaravan@hunter.
cuny.edu
All authors declare they have no conflicts of interest.
Funding for this article was provided by Blacksmith Institute, New York, NY.
Supplemental material attached to journal articles has not been edited,
and the authors take responsibility for the accuracy of all data.
http://dx.doi.org/10.1016/j.aogh.2014.08.002
Mexico, with its extensive lead ore deposits and widespread use of lead-glazed pottery, has a long and unique
history of lead exposure. Through technology and tradition, lead continues to threaten the health of millions
living in the country. In 1990, Mexico began phasing out
leaded gasoline and introduced unleaded fuels.1 Leaded
gasoline was completely phased out in Mexico by 1997.
The reduction of leaded gas has been linked to improved
health outcomes,2,3 however other means of exposure still
burden the people of Mexico.
Mining and secondary smelting also pose occupational and environmental health risks. Mexico is the fifth
largest producer of lead worldwide at 220,000 metric
tons in 2013 and has reserves of more than 5.6 million
tons.4 Three mining companies process lead ore from 13
mines operating in the states of Chihuahua, Coahuila,
270
Mexico Blood Lead Levels
Durango, Guererro, Hidalgo, San Luis Potosi, Sinaloa,
Sonora, and Zacatecas.5 Mexico is also one of the world’s
top 5 exporters of silver, often mined alongside lead ore,
and is a large producer, exporter, and recycler of lead-acid
storage batteries commonly used in automobiles and
lories.6,7
Leaded paints are a persisting health problem. The
Comex Group, North and Central America’s fourth
largest architectural paint manufacturer, and Sherwin
Williams S.A. de C.V., operate in Mexico’s large paint
and coatings production industry.8 A 2008 study
revealed that all tested samples of enamel paint contained lead concentrations greater than 90 ppm (the
regulatory limit in China and the United States); plastic
paints on average contained 6 ppm.9
A secondary source of general lead exposure in the
population is ceramic glazes used in traditional earthenware. Leaded glazes were first introduced by the Spanish
in the 16th century and are still used widely.10 Unfortunately, the vast majority of Mexican ceramic kilns are
wood-fired, as opposed to gas found elsewhere in the
world, and do not reach the fusing/sintering temperatures necessary to vitrify lead glazes to where they are
unleachable (temperature >1200 C). According to the
only census to date, at least 10,000 pottery workshops
use leaded glazes in wood-fired kilns.11 The problem is
compounded because many workshops are connected to
living and cooking areas, making area contamination
prevalent. These workers and their families are most
acutely at risk for lead poisoning and related illnesses. A
much broader exposure occurs with the use of leadglazed ceramics in the home, for meal preparation and
food storage. Lead can easily leach from the glaze into
food, where it is ingested. This situation is aggravated by
the Mexican diet, as lead is made more leachable by the
presence of heat or slight acidity such as that of lime
juice.12 Although lead-glazed pottery is abundant in
Mexico, poorly constructed, low-fired glazes are especially
a problem in rural areas where resources are scarce.
Low-income communities often are disproportionately
affected.
Measuring the scope of lead contamination in
Mexican communities is essential to fully characterize the
country’s burden of disease. This in turn helps define
interventions to mitigate exposures and alleviate adverse
health outcomes. In the United States, the National
Health and Nutrition Examination Survey (NHANES)
of the Centers for Disease Control and Prevention
regularly monitors blood lead levels (BLLs). Mexico has
no comparable program. In the absence of comprehensive data, we conducted an extensive historical review,
cataloging those studies that incorporate BLL testing in
Mexican communities. To our knowledge, this is the first
such literature survey of BLL tests in Mexico.
The literature on health effects from lead exposure is
extensive and definitive. Lead toxicity has been linked to
various cognitive impairments, lowered IQs, cardiovascular
effects, low birth weight, added economic costs, overall
diminished life expectancy, and possibly even increased
rates of violent crime.13-19 The second component of this
paper calculates the pediatric burden of disease from lead
using methods developed by the World Health Organization (WHO). Using values from the literature review, we
generated the attributable disability-adjusted life-years
(DALYs) for pediatric exposure to lead in Mexico. We also
estimated the effect on IQ, broadly providing a comparison
against the United States, a country that does not have
extensive exposure to lead-glazed pottery.
METHODS
Historical Review and Data Selection
As of January 2014, a PubMed search for lead AND levels
AND Mexico yielded 484 articles. Various permutations
were tested to ensure a large capture with these final search
terms selected. We reviewed articles in English and
Spanish, including only those most relevant to the study.
Our selection criteria retained 83 papers as eligible for
meta-analysis. A study was included in our analysis if the
authors 1) collected BLL data from populations within
Mexico; 2) upheld lead as a main focus; 3) included at
least 30 participants; and 4) reported BLL data from either
venous, capillary, or umbilical cord samples.
Studies ineligible for our meta-analysis met one or
more of the following exclusion criteria: 1) did not report
any BLL data and exclusively sampled from bone, organs, or other tissues; 2) lead was not the main focus of
the study; 3) BLL data were reported for Mexican populations existing outside of national Mexico; 4) the study
was not directly human health related; 5) the number of
study participants was less than 30; 6) the paper was out
of print and/or irretrievable; and 7) the study did not
contain a statistical mean or SD for the original data set.
Figure 1 summarizes the inclusion and exclusion criteria
for studies in our analyses. Exclusion criteria removed
82% of the studies from our initial search. Because many
of the publications we selected included multiple unique
samples (each with a unique estimated BLL), we hereafter
refer to these individual estimates as “samples.”
Subgroup Rationale
Mean and SD data from the 83 studies were separated
into subgroups for BLL analysis based on demographic
and sampling information. Means were collected based
on the initial year of data collection, not the study publication year. In cases where data was collected over a
period of time, we used the middle year. In instances
where no collection year was provided, we used the year
before article submission. The meta-analysis included
data for the following subgroups: sex (male and female),
age, urbanicity, workforce exposures, and blood sample
source (umbilical cord—hereinafter referred to as cord
samples—and venous/capillary; see Fig. 1).
271
Annals of Global Health
Among age groups, the categories were defined as
infant (<1 year), child and adolescent (1-14 years), and
adult (15þ years). Cord blood samples, presumably
taken at birth, were classified as infant along with any
venous or capillary blood samples taken within the participant’s first year of life. In cases when a study did
not meet our range or criteria, it was defined as “not
specified.”
The population density of the data collection area
was used to determine “urbanicity” for each study. We
used the US Census Bureau’s definition of an urbanized
area (a population density of 1,000 people per square
mile) to categorize each study.13 In many cases, this was a
fairly straightforward assessment. For instance, any study
where data were collected from Mexico City, the study
was immediately coded as “urban.” Conversely, in
obviously rural areas with very sparse populations (such
as agricultural areas), the studies were coded as “rural.”
For “suburban” areas or small cities where population
density could not be determined from a Web search, our
study coordinator in Mexico made the final decision.
This only applied to four cases.
Occupational exposure was defined as any cohort of
participants that was exposed to lead due to work (typically for 8 hours/day). According to Haz-Map from the
US National Institutes of Health (NIH), this includes
any worker involved in mining, smelting, manufacturing
with lead, ceramic or pottery workers, or painters using
leaded paint. Studies involving radiator repair workers
and lithographic print shop workers who were included
in this review were also considered to have occupational
exposure.
Blood sample source was determined by distinguishing cord blood samples from venous and capillary samples. As previously noted, cord blood samples
were grouped together with venous or capillary samples
taken from children under the age of 1 year in a general
“infant” category.
Figure 1. Summary of studies included and excluded in the literature review.
We excluded 82% of the articles reviewed from the literature.
years included (Fig. 2). To assess temporal trends in BLLs,
we created scatterplots of each study’s GM BLL against the
initial year of data collection (Fig. 3 and 4). These scatterplots included a LOESS (local regression) regression
Sex
Female 9.5 (7.7−11.4)
Male 18.5 (11.1−25.9)
Meta-analysis of BLLs
Due to the typically skewed distribution of blood lead data
we analyzed geometric mean (GM) data. In cases where
studies presented natural scale BLLs, we employed a previously described method to convert natural scale means
to geometric means.14 Overall and subgroup means were
calculated by weighting individual study results by their
inverse variance.15,16 We tested for heterogeneity of study
means through the Q statistic. If the Q statistic was statistically significant, we employed a random-effects model;
otherwise we employed fixed-effects model. When
analyzing subgroups (e.g., males or females), we provided
Q statistic results for each subgroup and report metaregression results to test whether subgroup results differ
with statistical significance.
Because of the large number of constituent studies in
the meta-analysis, we created a forest plot that illustrates
only the overall and subgroup summary results for all
Age
Infant 6.5 (5.5−7.6)
Child 9.7 (8.6−10.8)
Adult 10.9 (8.1−13.6)
Urbanicity
Urban 8.9 (8.1−9.6)
Rural 22.2 (13.5−31.0)
Workforce Exposure
Occupational 32.8 (20.2−45.5)
Non−Occupational 8.9 (8.2−9.5)
Blood Sample Source
Cord 6.4 (5.5−7.4)
Venous/Capillary 10.9 (9.4−12.5)
0
10
20
30
40
Geometric Means of Blood Lead with 95% CI's (ug/dL)
Figure 2. Forest plot of all available data (150 samples).
272
Mexico Blood Lead Levels
conducted with R ¼ 3.0.2 (R Foundation for Statistical
Computing, Vienna, Austria), using the “metafor” package. For all analyses, results were considered significant at
P < .05.
Figure 3. Scatterplot of all available data with smoothed regression lines for
urban and rural sample locations.*Dot size is proportionate to the inverse
variance of each sample’s estimate. The dotted-line intersecting this figure denotes the most recently available US average BLL of 1.2 mg/dL, given in geometric mean for a 95% CI. The shaded area between 1990 and 1997 indicates
the total period of the tetraethanol (leaded gasoline) phase out in Mexico.
line that was weighted by each study estimate’s inverse
variance. The US blood lead GM of 1.2 ug/dL from 2009
to 2010 was included in the plot (as a horizontal dotted
line), for visual comparison.17 A semitransparent rectangle
was added to highlight the entire period of leaded gasoline
phase out in Mexico (1990-1997). Our analysis was
Calculating Disability-Adjusted Life-Years and
IQ Impact. DALYs are a combination of 2 metrics,
namely years lived with disability (YLD) and years of life
lost (YLL) from a disease or health condition.18 For
lead in children, we focused on mild mental retardation
(MMR) as the main disability associated with increased
lead exposure. Because MMR is not inherently fatal, we
did not consider YLL in our analysis. To calculate YLD,
the duration of the disease in years was multiplied by a
disability weight (DW), which ranges from 0 to 1,
indicating severity. In YLD calculations, a DW of 0 indicates perfect health, whereas a DW of 1 would indicate a full year lost to death. By way of example, low
vision cataracts have a DW of 0.170 and untreated
asthma has a DW of 0.043. MMR due to lead has a
DW of 0.361.19
Two WHO DALY lead poisoning spreadsheets
were used to calculate the DALYs from lead exposure.
The first spreadsheet calculates the incidence of MMR
and requires entering an estimate for the GM BLL (with
a geometric SD). This spreadsheet calculates the number
of children above the MMR threshold of 70 IQ points
who would drop into the MMR range due to leadinduced cognitive impairment and presents the rate of
lead-induced MMR per 1000 population. It also incorporates a regional adjustment ratio for MMR since
Table 1. Details on Subgroups Within All 150 Samples
Subgroups Within Samples
Number of Samples, %
Sex
Female
61 (40.7)
Male
20 (13.3)
Not specified
Age
69 (46)
Infants
23 (15.3)
Children and adolescents
61 (40.7)
Adults
56 (37.3)
Not specified
10 (6.7)
Urbanicity
Urban
Rura
Not specified
132 (88)
18 (12)
0 (0)
Workforce exposures
Occupational
Nonoccupational
Figure 4. Scatterplot of all available data with smoothed regression lines for
occupation and nonoccupational exposure.*Dot size is proportionate to the inverse variance of each sample’s estimate. The dotted-line intersecting this figure
denotes the most recently available US average BLL of 1.2 mg/dL, given in geometric mean for a 95% CI. The shaded area between 1990 and 1997 indicates
the total period of the tetraethanol (leaded gasoline) phase out in Mexico.
Not specified
10 (6.7)
140 (93.3)
0 (0)
Blood sample source
Cord blood
Venous/capillary
Not specified
16 (10.7)
134 (89.3)
0 (0)
273
Annals of Global Health
Table 2. Subgroup Analysis of All 150 Samples
Estimated Subgroup Geometric
Subgroup
Q Statistic
Model Type
Mean in ug/dL (with 95% CI)
Sex (test for differences between subgroups: P ¼ .0009)
Female
Q ¼ 3862.80, df ¼ 60, P < 0.0001
Random-effects model
9.52 (7.67e11.36)
Male
Q ¼ 7867.66, df ¼ 19, P < 0.0001
Random-effects model
18.47 (11.05e25.89)
Age (test for differences between subgroups: P ¼ 0.05)
Prenatal/infant
Children and adolescents
Q ¼ 118.063, df ¼ 22, P < 0.0001
Q ¼ 2487.51, df ¼ 60, P < 0.0001
Random-effects model
Random-effects model
6.54 (5.46e7.61)
9.70 (8.64e10.77)
Adult
Q ¼ 5428.29, df ¼ 55, P < 0.0001
Random-effects model
10.87 (8.10e13.63)
Urbanicity (test for differences between subgroups: P < 0.0001)
Urban
Q ¼ 3898.77, df ¼ 131, P < 0.0001
Random-effects model
8.85 (8.09e9.62)
Rural
Q ¼ 5305.46, df ¼ 17, P < 0.0001
Random-effects model
22.24 (13.54e30.95)
Workforce exposures (test for differences between subgroups: P < 0.0001 )
Occupational
Q ¼ 3410.03, df ¼ 9, P < 0.0001
Random-effects model
Nonoccupational
Q ¼ 3624.74, df ¼ 139, P < 0.0001
Random-effects model
Blood sample source (test for differences between subgroups: P ¼ 0.04)
32.84 (20.19e45.50)
8.85 (8.17e9.52)
Cord blood
Q ¼ 42.19, df ¼ 15, P ¼ 0.0002
Random-effects model
6.45 (5.52e7.38)
Venous/capillary
Q ¼ 13046.88, df ¼ 133, P < 0.0001
Random-effects model
10.91 (9.36e12.47)
the incidence of noncongenital causes of MMR, such as
anemia and meningitis, varies in different regions of the
world. Additionally, the spreadsheet calculates the percentage of individuals with BLLs greater than 5 mg/dL
using the size of the population and mean BLL from the
population. This BLL was determined by the historical
review described earlier. The GM of 6 blood lead samples from studies in Mexico with data collection occurring in 2000 or later where only general community level
exposure to lead in children and adolescents could be
ascertained.
The second spreadsheet uses the incidence data and
various population values (e.g., age groups, total population) to calculate overall DALYs and DALYs per 1000
children associated with lead-induced MMR. There is no
DW associated with decrement of IQ points that does
not result in MMR. DALY calculations typically incorporate only the lead-induced cognitive impairment
resulting in MMR.20 A child who develops MMR from
lead at birth and has a life expectancy of 80 years is said
to have lived 28.9 YLDs because 80 years 0.361 equals
28.9 Although the calculation of DALYs from lead
typically involves calculating YLL, which represents the
premature mortality from a disease, from increases in
blood pressure and the resulting cardiovascular disease,
we did not include this component because the analysis
is limited to children younger than age 4 years. Therefore, our DALY calculations represent solely the
morbidity associated with lead exposure at these sites and
does not incorporate any premature mortality associated
with lead exposure.
Using the initial WHO spreadsheet, we calculated
the decrement of IQ points from the estimated BLLs
and DALYs resulting from incident cases of MMR.
This spreadsheet employed a previously developed model
(2005) that documented a reduction of IQ points equal
to e2.70 ln (concurrent BLL).21
Sensitivity Analysis
We performed a sensitivity analysis by calculating
DALYs using different discount rates and age weights.
Calculation of DALYs typically incorporates a 3% discount rate due to the societal preference of a healthy year
of life now versus in the future, and nonuniform age
weights due to the relative societal value of certain ages.22
The notation DALYs(r,K) signifies which discount rate (r)
and age weight (K) is used. We present our primary results in DALYs(3,1) including both the 3% discount rate
and the full age weight. We also present DALYs(3,0) with
only the 3% discount rate and DALYs(0,0) without any
weighting.
RESULTS
Analysis of All Available Data
The final data set reviewed for meta-analyses included 83
studies with participants from throughout Mexico, and
resulted in 150 data samples. The data samples ranged
from years 1978 to 2010, wherein years were used to
define the period of sample collection and not article
publication year.
The BLL GMs for various subgroups are shown in
Figure 2 for comparison. The dotted line indicates the
current reference level for elevated blood lead as described
by the CDC of 5 ug/dL.23 No international reference
dose currently exists, although WHO and others use the
CDC standard. GMs by subgroup are as follows: sex
(males 18.5 mg/dL, females 9.5 mg/dL); age (infants 6.5
mg/dL, children and adolescents 9.7 mg/dL, adults 10.9
274
Mexico Blood Lead Levels
Table 3. Geometric Means from 9 “General Exposure” Samples Collected from 2000 Onward
Year of Initial Data Collection
Primary Author
2000
Gomaa†
2000
Afeiche M†
2000
Study Sample Size
Geometric Mean (with 95% CI)*
100
5.97 (5.44e6.56)
1000
3.02 (2.9e3.15)
Torres Alanis†
207
10.55 (9.97e11.16)
2001
Lamadrid-Figueroa
207
5.22 (4.81e5.66)
2002
Ettinger
670
3.95 (3.94e3.96)
2002
2004
Vanitha Janakiraman
Téllez-Rojo MM†
31
566
6.04 (4.72e7.72)
3.97 (3.79e4.16)
2004
Téllez-Rojo MM†
752
4.71 (4.5e4.93)
2004
Téllez-Rojo MM†
583
4.66 (4.44e4.88)
Overall results:
Q statistic: Q ¼ 30.69, df ¼ 8, P ¼ 0.002.
Estimated overall geometric mean (with 95% CI) from random-effects model: 5.36 (3.90e6.82) ug/dL.
*95% CI’s calculated from geometric standard deviations provided in manuscript.
†
Pediatric populations included in subset meta-analysis.
mg/dL); urbanicity (urban 8.9 ug/dL, rural 22.2 ug/dL);
workforce exposure (occupational 32.8 ug/dL, nonoccupational 8.9 ug/dL); blood sample source (umbilical cord
6.4 mg/dL, venous/capillary 10.9 mg/dL).
Figures 3 and 4 depict scatterplots with LOESS
regression lines for Mexican rural populations (GM
22.2 mg/dL) and urban populations (GM 8.9 mg/dL),
and occupational exposure (GM 32.8 mg/dL) and
nonoccupational exposure (GM 8.9 mg/dL), respectively.
A dotted line intersecting Figure 2 and 3 references, for
comparison, the most recently measured average BLL for
the United States at 1.2 mg/dL.24 This value is given in
GM with 95% confidence interval. A marker for the
years 1990 to 1997 bisects Figures 2 and 3, indicating
the total period of the tetraethanol (leaded gasoline)
phase out in Mexico.
We provide further information regarding the 150point data set in Tables 1 and 2. Here we show the
composition of group populations in relation to the
entire data set; therefore each subgroup totals 150 samples. More studies sampled females than males (61 of the
150 studies, 40.7% vs male 13.3% and not specified
46%). Other trends in study samples featured urban
populations (88% of studies) for nonoccupationally
exposed groups (93.9% of studies) tested through
venous/capillary sampling (89.3% of studies).
Analysis of Studies Since 2000
We analyzed recent samples (i.e., data collected since
2000) for relative comparison with modern lead toxicity
standards. In all, 36 samples (GMs) met the inclusion
criterion. To reflect the outcomes of general exposure,
studies in which participants were at risk for occupational exposure or increased community exposures were
removed from the analysis. Therefore, our general
exposure analysis does not include studies in which
participants faced possible increased exposure due to
living near smelters, used lead-acid battery facilities,
mine-tailing zones, mining regions, metal foundries,
radiator repair shops, lithographic print shops, metal
recycling facilities, or industrial zones. Nine samples were
included in the final meta-analysis (N ¼ 4,116). The GM
for BLLs in included studies since 2000, after weighting
individual study results by their inverse variance, is 5.36
mg/dL. This result slightly exceeds the 5 mg/dL US
Table 4. Subgroup Analysis of 9 Samples
Estimated Subgroup Geometric
Subgroup
Q Statistic
Model Type
Mean in ug/dL (with 95% CI)
Age (test for differences between subgroups: P ¼ 0.56)
Prenatal/infant
Q ¼ 1.61, df ¼ 3, P ¼ 0.66
Fixed-effects model
4.86 (3.71e6.01)
Child and adolescent
Adult
Q ¼ 20.43, df ¼ 1, P < 0.0001
Q ¼ 1.77, df ¼ 2, P ¼ 0.41
Random-effects model
Fixed-effects model
6.81 (0e14.18)
4.89 (3.60e6.18)
Blood sample source (test for differences between subgroups: p ¼ 0.98)
Cord blood
Q ¼ 0.66, df ¼ 1, P ¼ 0.42
Fixed-effects model
5.35 (3.76e6.94)
Venous/capillary
Q ¼ 30.02, df ¼ 6, v < 0.0001
Random-effects model
5.37 (3.49e7.25)
Overall results among 6 “general exposure” samples collected from 2000 onward that only included children, adolescents, or infants:
Q statistic: Q ¼ 27.95, df ¼ 5 P < 0.0001.
Estimated overall geometric mean (with 95% CI) from random-effects model: 5.52 (3.35e7.68) ug/dL, Geometric SD: 2.70.
275
Annals of Global Health
Table 5. Inputs for Calculation of Lead Poisoning MMR
DALYs
Population (N)
117,886,000
Population (0-4 y)
11,573,000
Regional adjustment ratio
2.70
Incidence of MMR (0-4, per 1000)
5.98
Blood lead level Geometric mean
BLLeSD
5.52
2.70
DALY, disability-adjusted life-year; MMR, minor mental
retardation.
recommended intervention level. Details are shown in
Table 3.
Analysis of Subgroups Since 2000
Due to the limited sample size of studies since 2000, only
the age- and blood-sample category subgroups were
analyzed under the Q statistic, viewable in Table 4. Each
of these studies analyzes urban populations with
nonoccupational exposures to lead. Studies since 2000
sample infants (4 samples, N ¼ 2,001), children and
adolescents (2 samples, N ¼ 1,207), and adults (3
samples, N ¼ 908). Estimated subgroup GMs were as
follows: infant, 4.86 ug/dL; child and adolescent, 6.81
ug/dL; adult 4.89 ug/dL; cord blood, 5.35 ug/dL; and
venous/capillary blood, 5.37 ug/dL.
IQ Deficits Associated with MMR from
Lead
Given the GM of 5.52 ug/dL (calculated from the 6
general exposure samples collected from 2000 onward
in infants, children, and adolescents), IQ deficits
associated with MMR from lead exposure were calculated as well. In a random sample of 1000 Mexican
children, 156 (>15%) are expected to have a reduction
of more than 5 IQ points due to lead exposure.
Another 221 (>20%) are expected to have a reduction
of more than 2 IQ points due to lead exposure. In its
most severe cases of generalized exposure, lead is expected to cause 44 of 1000 children to have a decrease
of nearly 7 IQ points.
BLLs and DALYs
After completing all calculations within the WHO
spreadsheets for DALYs associated with lead poisoning
and MMR, it was determined that the incidence rate of
MMR due to lead in children aged 0 to 4 in Mexico is
5.98 per 1000 children, viewable in Table 5. The GM of
5.52 ug/dL was determined by calculating a geometric
mean from 6 general exposure samples collected from
2000 onward that only included children and infants.
Including a 3% discount rate and the full age weight, this
amounts to 820,548 DALYs from lead-induced MMR
for children aged 0 to 4 years in Mexico. Applying a 3%
discount rate and no age weight, the amount is 749,839
DALYs for the same group. Without either a discount
rate or age weight, this number increases to 1,920,962
DALYs. Adjustment of discounting and age weights offers a range of estimates while reflecting the relative social
values of age groups.25 When compared with other
sources of morbidity and mortality for children in
Mexico, the DALYs associated with lead-induced MMR
are extremely significant.
DISCUSSION
Because our meta-analysis summary represents GMs from
all data within the period between 1978 and 2010, we
consider this a comprehensive “historical” review. Analyses
of subgroups were also essential, as some populations are
predisposed to practices that lead to greater lead exposure.
Our results fall in line with the expected trends in lead
exposure. Furthermore, blood was chosen as the ideal
sample medium for measuring lead levels most accurately.
Bone also sequesters lead; however the half-life of lead in
bone is approximately 20 years.26 Therefore, bone does not
accurately depict the level of existing lead exposure. However, high bone lead concentrations can continue to leach
Pb into the blood where systemic effects can continue.
As anticipated, the GM for BLLs in males was
significantly higher than the average for females; the
overall average BLL for males was about 1.9 times higher
than that of females. Although women make up the
majority of lead-glazed pottery workers, smelting and
recycling are primarily male-dominated professions in
Mexico. These occupational exposures are the likely
cause for the elevated levels in males.
The meta-analysis showed average BLLs increased
with age. This is not surprising given the half-life of lead
in blood is approximately 1 month (28-36 days).27 Lead
exposure from pottery is ubiquitous throughout the
country, especially in poorer communities. Occupational
exposures and trace levels from pre-1990 leaded gasoline
are other possible exposure sources.
The use of lead-glazed pottery is more common in rural
areas than in urban zones.28 Additionally, lead smelting
activity occurs at much higher frequencies in rural areas.
Therefore, the significantly higher GMs for rural BLLs
(roughly 2.5 times higher than in urban areas; overall 22.24
and 8.85 ug/dL, respectively) met our general expectations.
Workforce Exposure Scatterplot
In our analysis of workforce exposure, every study was
classified as occupational or nonoccupational as previously
described. The smoothed regression line for nonoccupational studies shows a modest and relatively steady
decrease since our first data collection year of 1978.
Occupational, on the other hand, appears to shows a
sharp decline. This can be explained by a number of
factors. First, there are several outliers in the beginning of
our data collection period from the 1970s and 1980s that
are necessary to show the full picture of our studies that
276
Mexico Blood Lead Levels
Table 6. Relative Ranking of Childhood DALYs in
Mexico
Cause
Lead poisoning
I. Communicable, maternal,
DALY
820,000
1,890, 000
perinatal, and nutritional conditions
Infectious and parasitic diseases
Tuberculosis
sexually transmitted diseases
340,000
3000
8,000
excluding HIV
Diarrheal diseases
179,000
Hepatitis
5000
Malaria
<1000
Respiratory infection
301,000
Perinatal conditions (low birth
weight, neonatal inf)
1,059,000
Nutritional deficiency (protein,
189,000
iodine, vitamin A, iron)
II. Noncommunicable diseases
Malignant neoplasms
Diabetes mellitus
Neuropsychiatric conditions
2,260,000
79,000
2000
715,000
(depression, bipolar, epilepsy,
alcohol use, multiple sclerosis,
was expected that urban blood lead samples would be
lower than rural samples.
Recent Sample Analysis
GMs analyzed after 2000 with only community-level
exposure yielded an overall GM of 5.36 ug/dL. Furthermore, for 6 general exposure samples in infants, children,
and adolescents, the GM was calculated to be 5.52 ug/dL.
Although the data included in the 6 samples that were
used to calculate our final GM were not explicitly nationally representative, they are the best representation of
general exposure lead in Mexican infants, children, and
adolescents based on the recent literature. If this sample is
any indication of the overall BLL mean, it means that
children in Mexico have BLLs that are on average 4.7
times higher than their US counterparts (recent NHANES
studies have shown the levels to be around 1.17 ug/dL).29
The upper value of the reference range is defined as
97.5% of the sampled population (1-5-year-old children in
the United States). In 2012, this upper limit was estimated
to be 5 ug/dL. A calculated GM of 5.52 means that nearly
half of the population of Mexican children is above a value
that has been demonstrated to negatively affect children’s
intelligence and behavior. To compare, this is only 2.5%
of children in the United States.
drug use, migraine)
Cardiovascular diseases
37,000
DALYs and IQ Deficits
Asthma
185,000
Genitourinary diseases
39,000
Oral conditions (dental caries,
116,000
Our analysis indicates that more than 820,000 DALYs
result from lead-induced MMR in Mexican children (after
accounting for a 3% discount rate and full age weight).
This represents an enormous source of disability in
Mexico. As Table 6 shows, lead in Mexico causes more
DALYs than neuropsychiatric conditions (715,000), infectious and parasitic diseases (340,000), respiratory infections
(301,000), nutritional deficiency (189,000), and all injuries
(473,000), among other illnesses. Lead exposure also likely
results in significant IQ decrements that do not result in
MMR and are therefore not captured by the DALY metric.
We estimate that 15% of the population aged 0 to 4 will
experience an estimated IQ decrement of more than 5
points from lead exposure, resulting in more than 8.6
million lost IQ points in Mexico for this population.
periodontal disease)
III. Injuries
Road traffic accidents
473,000
121,000
Falls
51,000
Drownings
36,000
Other unintentional injuries
211,000
Intentional injuries (self-inflicted,
37,000
violence, war)
met inclusion criteria. Additionally, only 10 samples
reviewed could be classified as solely “occupational.”
Therefore, this precipitous decrease in the average occupational BLLs should be reviewed cautiously.
Urbanicity Scatterplot
Figure 3 uses all 150 samples included in our review and
every study was classified as urban or rural. The biggest
sources of lead exposure in rural areas of Mexico are
believed to be from the use of lead-glaze pottery, smelting,
and pottery making itself. In urban areas, pottery making
and residual dust from leaded gasoline have historically
been the biggest sources, although lead-glazed ceramics
are most likely the most significant source. In rural areas,
many studies reviewed focused on subpopulations of
increased exposure, such as artisanal communities,
where BLLs in children are extremely high. Therefore, it
Limitations
There are several limitations in our analysis. Most significant
among them is our reliance on a limited number of reports.
In the absence of a more comprehensive dataset, we used a
finite number of studies in our approach and developed
estimates that we then applied to the entire Mexican population. We believe this approach, used elsewhere, is useful
and illuminating, given that a large, population-based sampling of blood lead was never conducted. Similarly, many of
the studies we used focused on populations known to be
exposed to higher levels of environmental lead, meaning
they were targeted studies and not always random population
samplings. We articulated this difference by disaggregating
general exposure from elevated exposure (Table 5); however,
277
Annals of Global Health
the net number of general exposure population studies was
limited. Finally, we assumed that the primary risk factor for
elevated BLLs in the general population of Mexico was
exposure to leaded ceramic glazes. We accounted for this by
indicating the relative decline in leaded gasoline and lack of
other known exposure pathways. This link could be better
verified through further study.
CONCLUSION
Data compiled from 83 studies, representing 150 data
points from more than 50,000 study participants,
demonstrated a trend of elevated BLLs in Mexico between 1978 and 2010. The public health implications of
lead contact even at low levels are predicted to have severe and chronic effects with continued exposure.
Although comprehensive national data are unavailable
for average population BLLs in Mexico, recent data
(studies since 2000) suggests lead exposure remains a
national-level public health concern.
Although blood lead has decreased significantly in
Mexico over the past 35 years, it remains significantly
elevated. The post-leaded gasoline average BLL is still
more than 4.5 times higher than the level in the US (5.52
vs 1.2 ug/dL). We estimate that this will result in 820,000
DALYs and 8.6 million fewer IQ points from leadinduced MMR in Mexican children aged 0 to 4 alone.
In 2013, the Mexican Secretariat of the Interior
officially censured the use of lead in ceramics.30 Affordable and effective alternative glazes exist, although uptake
by artisanal cermacists has been slow. Barriers are both
economic and cutural in nature. A more aggressive
approach focused on both the consumer and producer
might begin to make progress on this severe and insidious
risk to the health of the Mexican population.
Other major sources of lead—mining, smelting, and
paints—compound exposure rates and health outcomes.
A nationwide public health program is necessary to
monitor lead levels in children that can identify hot spots
and other sources of exposure for the general population,
and offer intervention where needed.
References
1. Environmental Protection Agency. Implementer’s Guide to Phasing
Out Lead In Gasoline. 1999. Available at: http://www.epa.gov/
international/air/pdf/EPA_phase_out.pdf. Accessed April 24, 2014.
2. World Health Organization. Childhood lead poisoning. 2006. Available at: http://www.who.int/ceh/publications/leadguidance.pdf.
Accessed April 3, 2014.
3. Schnaas L, Rothenberg S, Flores M, et al. Blood lead secular trend in
a cohort of children in Mexico City (1987e2002). Environ Health
Perspect 2004;12:1110e5.
4. US Geological Survey. Mineral commodity summary. 2014. Available
at: http://minerals.usgs.gov/minerals/pubs/country/2011/myb3-2011mx.pdf. Accessed February 12, 2014.
5. Perez AA, US Geological Survey. The mineral industry of Mexico 2011
minerals yearbook: Mexico. 2011. Available at: http://minerals.usgs.
gov/minerals/pubs/mcs/2014/mcs2014.pdf. Accessed July 2, 2011.
6. Guberman D. 2011 Minerals yearbook: lead advance release US geological rurvey. 2013. Available at: http://minerals.usgs.gov/minerals/pubs/
commodity/lead/myb1-2011-lead.pdf. Accessed May 7, 2013.
7. USGS. Mineral commodity summaries Reston, Virginia. 2013. Available at: http://minerals.usgs.gov/minerals/pubs/mcs/2013/mcs2013.
pdf. Accessed May 7, 2014.
8. IPEN. Lead in new decorative paints. 2009. Available at: http://ipen.
org/sites/default/files/documents/global_paintstudy-en.pdf. Accessed
July 22, 2014.
9. IPEN. Global study to determine lead in Decorative paints in 10
countries. 2009. Available at: http://ipen.org/documents/globalstudy-determine-lead-new-decorative-paints-10-countries. Accessed
July 22 2014.
10. Charleton TH. Contemporary Central Mexican ceramics: a view from
the past. Man 1976;11:517e25.
11. FONART. Cómo detectar la presencia de plomo en cazuelas, olla, platos
y jarros de barro esmaltado? 2009. Available at:http://alfareria.org/sites/
default/files/images/ManualPruebas.pdf. Accessed April 17, 2014.
12. Feldman N, Lamp C, Craigmill A. Lead leaching in ceramics difficult
to predict. Calif Agriculture 1999;53:20e3.
13. U.S. Department of Health and Human Services. 2013. Health information technology: how is rural defined? Available at: http://
www.hrsa.gov/healthit/toolbox/RuralHealthITtoolbox/Introduction/
defined.html.
14. Higgins JPT, White IR, Anzures-Cabrera J. Meta-analysis of skewed
data: combining results reported on log-transformed or raw scales.
Statist Med 2008;27:6072e92.
15. Lipsey MW, Wilson D. Practical Meta-Analysis. Thousand Oaks, CA:
Sage Publications; 2001.
16. Whitehead A. Meta-analysis of Controlled Clinical Trials. Chichester,
UK: Wiley; 2002.
17. Centers for Disease Control and Prevention. National report on human
exposure to environmental chemicals (updated tables, September 2013).
2013. Available at: http://www.cdc.gov/exposurereport/. Accessed April
22, 2014.
18. World Health Organization. Global health risks: mortality and burden
of disease attributable to selected major risks. Geneva: WHO; 2009.
19. World Health Organization. Global burden of disease 2004 update:
disability weights for diseases and conditions. 2004. Available at:
http://www.who.int/healthinfo/global_burden_disease/GBD2004_
DisabilityWeights.pdf.
20. Schwartz J. Low-level lead exposure and children’s IQ: a meta-analysis
and search for a threshold. Environ Res 1994;65:42e55.
21. Lanphear BP, Hornung R, Khoury J, et al. Low-level environmental
lead exposure and children’s intellectual function: an international
pooled analysis. Environ Health Perspect 2005;113:894.
22. World Health Organization. Global health risks: mortality and burden
of disease attributable to selected major risks. 2009. Available at:
http://www.who.int/healthinfo/global_burden_disease/GlobalHealthRisks_
report_full.pdf.
23. The 5 ug/dL reference level is based on the 97.5th percentile of the
National Health and Nutrition Examination Survey, or NHANES, blood
lead distribution in children. CDC updates the reference value every
four years using the two most recent NHANES surveys. Available at:
http://www.cdc.gov/nceh/lead/ACCLPP/blood_lead_levels.htm.
24. Agency for Toxic Substances & Disease Registry. Lead Toxicity: What is
the Biological Fate of Lead? August 20, 2010. Avilable at: http://www.
atsdr.cdc.gov/csem/csem.asp?csem¼7&po¼9. Accessed May 6, 2014.
25. Mathers CD, Salomon JA, Ezzati M, Begg S, Vander Hoorn S,
Lopez AD. Sensitivity and uncertainty analyses for burden of disease
and risk factors estimates. In: Lopez AD, Mathers CD, Ezzati M, et al.,
eds. Global Burden of Disease and Risk Factors. Washington DC:
World Bank; 2006:399e426.
26. Rabinowitz MB. Toxicokinetics of bone lead. Environ Health Perspect
1991;91:33e7.
27. Agency for Toxic Substances & Disease Registry. Lead toxicity: what is
the biological fate of lead? Available at: http://www.atsdr.cdc.gov/
csem/csem.asp?csem¼7&po¼9. Accessed May 6, 2014.
28. Lilia AA, Badillo F. Environmental Lead in Mexico: Reviews of Environmental Contamination and Toxicology. New York: Springer; 1991:1e49.
29. US Environmental Protection Agency. Blood lead. Available at: http://
cfpub.epa.gov/roe/indicator_pdf.cfm?i¼63. Accessed April 24, 2014.
30. SEGOB. Limitaciones y especificaciones sanitarias para el uso de los
compuestos de plomo. 2014. Available at: http://www.dof.gob.mx/
nota_detalle.php?codigo¼5343154&fecha¼02/05/2014.